Dynamic, Multi-modal Imaging of Additive Manufacturing

Our phase methods use apertured masks to make x-ray systems sensitive to phase effects: these are extracted by analysing the distortions in the beamlets created by the mask. If the detector resolution is smaller than the beamlets, the sample’s attenuation, phase and scatter signals can be simultaneously extracted from a single frame. However, the resolution will be limited by the spacing between the apertures. To overcome this, we have collected images while continuously scanning the mask in front of the sample. The latter was an aluminium-based powder bed, molten in real time by a high-powered laser. The method produces three complementary and automatically registered image sequences, basically attenuation, phase and scatter videos of the powder as it melts. Not only did this show that the formation of liquid droplets is detected much earlier in the phase and scatter videos than in the attenuation ones; it also allowed tracking the movement of the powder particles leading to the identification of “accumulation zones” where the droplets subsequently form. In other words, the method allows predicting melting locations before the melting actually begins. Calibration of the scatter signal also enabled estimating the local thickness of unfused powder, i.e. the extraction of a degree of 3D information from a 2D image.

Dynamic sequence showing how the formation of liquid droplets becomes apparent much earlier in the phase (middle row) and dark-field (bottom row) images compared to conventional x-ray attenuation (top row). Time (in milliseconds) from when the laser hits the powder bed is reported in all images.